PSI - Issue 38

Tiago Werner et al. / Procedia Structural Integrity 38 (2022) 300–308 T. Werner/ Structural Integrity Procedia 00 (2021) 000 – 000

304

5

1 mm extension, although this did not comply with the recommendation of the standards. According to the prescriptions in the standards, a minimum pre-crack corresponding to the sum of the notch length and an additional 1 mm should have been produced in order to avoid any influence that the shape of the crack front or the loading history during pre-cracking might have on the calibration of the analytical expressions defining the K value. The two main reasons for refuting this requirement were a) to have a sufficiently large range of experimental values to compare results throughout about two orders of magnitude in terms of d a /d N , and b) to avoid an excessively small value of the force amplitude during testing that would lead to anomalies or inaccuracies in the machine's performance. It should be noted that, during the miniature specimen tests, the maximum load amplitude value was in all cases below 200 N, even as low as 37 N. Therefore, maintaining a moderate variation in the F max values during each test was a challenge when working with a servo-hydraulic machine with 10 kN load capacity. Nevertheless, it was always possible to keep the force peak values around ±1% of the nominal applied value, which conforms well to the demands of the standards (±2%). The fracture surface in each test has been analyzed using a Keyence VHX-2000 optical microscope. In fact, the value of a 0 in Eq. (1) can be measured only at the surface before starting the test. Therefore, a post-mortem analysis of the specimens under the microscope is necessary to correct the a 0 value originally evaluated. Likewise, the average value of the crack size at the end of the test is determined as well. Based on these optical measurements, the value of y in Eq. (1) can be optimized and the crack length values corrected over the entire crack propagation range. Figure 2 shows an example of a result when correction is evaluated in terms of crack size versus the number of cycles. Although the deviations when corrections are completed are generally not significant, it is a good practice to minimize any source of uncertainty other than that arising from the inherent scatter in the material response. When such a correction is applied, the driving force values will increase or decrease slightly, depending on the direction of the compensation. In addition, the small effects of minimal misalignments ( ≤ 200  m) between the top roller and the crack plane in the case of the SEN B3 configuration, which are unavoidable, were also taken into account.

Fig. 2. Example of crack length correction.

3. Experimental results Each test has been extended to the limit that would allow large-scale yielding effects to be avoided. This limit is defined by the test configuration and the material yield strength ( R p0.2 ), where 960 MPa is the characteristic value for the steel under investigation. Eq. (2) establishes the maximum crack size for the validity of the linear elastic Δ K as a driving force in the case of bending specimens (according to ISO 12108). The λ -value is dependent on the distance between supports of each test configuration. ≤ − ( 3 ∙ ∙ 2 ∙ ∙ p0,2 ) 1/2 (2)